"We have an ocean of opportunity and we are fishing in a pond’s worth of it. We need more fishermen and women – and boats."
MiQ’s US CEO, Sean Reardon, was speaking at a lunch debate about connecting data to improve marketing messages, moderated by Campaign US’ associate editor, Oliver McAteer.
The table of media agencies and brands stressed that the move from pond to ocean could only be made through better application of data and AI – and more simplified, efficient ways of working.
Give the people what they want
"Consumers are literally telling us exactly what they want," said Joe Reinstein, global growth officer at Performics. "But we have to do a better job of marrying all of this customer and audience information with the digital signal data."
He said that the focus has been too much on first- and third-party data – metrics such as device and geography.
"Who you are is less important than what you want," he said. Marketers must marry up the information about the consumers themselves with information that segments them by intent.
"If they want gluten free beer it doesn’t matter what life stage they are in. We need to be able to deliver against that need without being stuck in one spectrum of a segmentation."
He added that marketers need to be more nimble in responding to this data to deliver a content experience that is aligned with consumer intent – in real time.
Companies need to work out their data strategy and ensure they have the resources they need in house to manage it, said Aaron Smith, Wavemaker’s chief client officer. Smith said he knows CPG companies who have 70 million visitors to their site per year and aren’t doing anything with that data.
Joining the dots
Data unification has to transcend martech and ad tech, argued Anna Percy-Dove, EVP strategy at FCB/SIX. She said: "We must come together to design journeys that are people-centred not digital-focused."
MediaCom’s global director of innovation, Liam Brennan, said that CPG brands are beginning to understand that they need to connect different sources of data across organisations – shopper teams together with brand marketing teams, media teams talking to sales teams. Connecting data to the advertising is easier said than done, he said, but it requires full end-to-end conversion.
Anheuser-Busch is heading towards more micro targeting and multiple segmentation mapping, said Samrat Saran, Anheuser-Busch’s senior director, brand insights. To do this, they have made ecosystem planning "much simpler" creating more direct links between different types of data.
"In 2019 there is a lot more symbiosis and more constructive partnership conversations than three years ago, when there were more arguments about who owns what," said Evan Hanlon, chief strategy officer US at GroupM.
Smith added that the industry is "maturing" when it comes to using consumer data – increasing restrictions on how companies use data are beneficial for both consumers are also "wising up" about their data – "which is a good thing for creating stronger relationships with brands."
More intelligent AI
McAteer asked how marketers were balancing AI with human intelligence, and the group established that they work best when they are deployed together, working symbiotically.
"The idea that AI is going to solve everything is hilarious to me," said Reardon. "AI is going to help create better efficacy and help rational decision making. Marketing has forever been a balance of emotional and rational."
"Humans work first and foremost on the emotional part of the brain," said Percy-Dove. "Making decisions based on rational fact is hard."
She said that AI is the only way to be scalable, and humans must be the checks and balances: "You can’t let it run on its own."
But Cartier Stennis, US head of market research at Twitter, pointed out that machine learning and AI are "really good at prediction, but not particularly great at interpretation".
"It’s hard to understand why the AI made a decision, even it’s its the right one," he said. "The way it learns is dictated by humans."
He used the example of targeting users in the US on Twitter based on language. If you are targeting for Spanish speakers, one way would be to simply ask people whether or not they speak Spanish. Or you can ask the more nuanced question of is it your dominant language? How strongly do you speak Spanish? Is it your preferred language?
"Having that nuance is going to help the algorithms perform a lot better," said Stennis. "The only way you can do that is to have the human element to think about what you are really trying to accomplish and how you are best going to enable the system to learn how to do the right thing."
AI has the potential to find the needle in the haystack [of data] that humans can’t, said Brennan – it’s all about optimising faster and smarter.
Balancing creativity and data
"Tech won’t save us," warned Hanlon. "It requires creativity and ideas."
But data can also inspire creativity, said Percy-Dove: "The conversation that data is killing creativity – I totally dispute that."
She added that there is a lot of focus on the media side of data but there can be "amazing magic" when media insights creative insights and data all come together. "The best work I’ve seen is when there’s a tight bond between creative and media agencies."
In Percy-Dove’s opinion clients under-use data.
"Where we’ve had success is where we have used media data, looking at search behaviors which creative agencies might not necessarily look at."
Tracking this data for an automotive brand, Percy-Dove could see where people were comparing four-wheel drives on two different brands, comparing safety rates, or trying to find out how comfortable that seventh car seat might be. "We were able to design a targeted creative journey using those data signals."
Smith from Wavemaker talked about his work with L’Oreal for their True Match range. To highlight True Match’s big range of 23 different colors they created a multi segmentation approach. They looked at data to understand the different segments they wanted to target – looking into people’s particular interests to understand the stories around their lives. They then used influencers of various ethnicities to tell different stories around the product, targeting them towards the most relevant people within those segments. "We overlaid data with good creative storytelling," said Smith.
Reardon said that this example shows the potential of data to be applied creatively. One thing that has been overstated, he said, is that Dynamic Creative Optimisation is going to change things overnight.
"That story gives me great hope for where we are making progress," he said. "We need to see that being repeatable across engagements and clients."
Smith added a warning that the media industry has almost shifted too far to the data side.
"We’re more analytically skewed and there isn’t enough creative force to match the data," he said. "How do we make sure the creatives and data experts are working together from start to finish? There are opportunities to improve that. It feels like there is this balance of data and creativity resources is critical."
Saran said that one of the biggest gaps – especially in CPG – is that the tools used to create the creative aren’t the same ones used to measure the effectiveness of the creative.
But it’s very early days, he said. "We are still trying to stitch the material together to make a cloth we can wear – then we can decide the style of the cloth… I hope to see a universe where everyone involved can sit in one room and work as a team, rather than the brand in the middle trying to deploy."